Environ. Sci. Technol. 1998, 32, 1656-1661
Biotransformation versus Bioaccumulation: Sources of Methyl Sulfone PCB and 4,4′-DDE Metabolites in the Polar Bear Food Chain R O B E R T J . L E T C H E R , * ,†,‡ R O S S J . N O R S T R O M , †,§ A N D D E R E K C . G . M U I R |,⊥ Centre for Analytical and Environmental Chemistry, Department of Chemistry, Carleton University, Colonel By Drive, Ottawa, Ontario, K1S 5B6 Canada, Environment Canada, Canadian Wildlife Service, National Wildlife Research Centre, 100 Gamelin Boulevard, Hull, Que´bec, K1A 0H3 Canada, Freshwater Institute, Department of Fisheries and Oceans, 501 University Crescent, Winnipeg, Manitoba, R3T 2N6 Canada
In the polar bear food chain from the Canadian Arctic, methyl sulfone (MeSO2-) PCBs and 4,4′-DDE were below detection in arctic cod ( 0.05) difference between the levels for ringed seal males and females. f There were two pools, each containing nine individual cod. Levels for both pools are shown.
°C or lower until further use. The ringed seals were adults (5-8 years of age), except for two juvenile females (85% and ca. 90%, respectively. All concentrations were normalized to the extractable lipid. The PCB/DDT concentrations were recovery-corrected using the average recovery of the five higher chlorinated 13C12-labeled PCB standards.
Assessment of Bioaccumulation and Biotransformation. The recalcitrant congeners CB153 (2,2′,4,4′,5,5′) and CB180 (2,2′,3,4,4′,5,5′) are resistant to biotransformation in many organisms (4, 15). They possess chlorine substitution patterns that are not favored by CYP1A- and CYP2B-type enzymes because of the lack of adjacent, chlorine-unsubstituted carbons (19). Analyte concentrations were normalized to CB153 (C′) to minimize the variation among individuals and species (19, 20). For example, C′x ) Cx/C153 and C′mx ) Cmx/C153, where x denotes the PCB congener and mx denotes the MeSO2-PCB congener of PCB x. Uptake from prey of most of the PCBs (except octachloroand nonachloro-PCBs), 4,4′-DDE, 4,4′-DDT, and their MeSO2 metabolites is likely to be similar for a given species because these compounds have log Kow values of 4-7 (21). Log Kow values for several trichloro- to heptachloro-MeSO2-PCBs and 3-MeSO2-4,4′-DDE were computer modeled (Advanced Chemistry Development (ACD) Labs, Toronto, Canada) and ranged from 3.82 to 6.51. Therefore, BF′ (i.e., ) C′predator/ C′prey) is a good measure of bioaccumulation. Assuming that CB153 represents the maximum bioaccumulation potential for slowly metabolized lipophilic compounds, BF′ for such compounds should be ≈1.0. A BF′ of 1.0. The degree to which definitive conclusions can be made on the relative importance of these two processes from ratios of CB153normalized concentrations depends on several factors. The following three general classes can be defined for bioaccumulation of MeSO2-PCB metabolites from prey to predator and can be used for any metabolite with formation potential from remaining precursor compounds that bioaccumulate in a predator. Class I. The precursor PCB congener is not present in the prey, but the metabolite is found in the prey and predator. Therefore, C′mx(seal)/C′mx(cod) and C′mx(bear)/C′mx(seal) would be the true bioaccumulation factors, BF′mx, of the metabolite VOL. 32, NO. 11, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1657
relative to CB153 in seal and bear, respectively. If there are sufficient numbers of MeSO2-PCBs for which this class applies, it may be possible to develop a general relationship for prediction of BF′mx for those congeners that are metabolized slowly. Class II. The metabolite is not present in the prey but is present in the predator. C′mx in the predator is entirely due to metabolic formation from precursor PCB in the prey, and BF′mx has no meaning. Thus, the x fraction metabolized to mx and retained in the predator is MF′mx ) C′mx/M′x. Class III. Both metabolite and precursor PCB are found in the prey. If the apparent metabolite bioaccumulation factor, AF′ ) C′mx(predator)/C′mx(prey), in the predator is higher than the BF′mx predicted from class I, both bioaccumulation and metabolic formation are probably contributing to the observed metabolite levels in the predator, but bioaccumulation alone may be occurring up to AF′ ) 1.0. Assuming that BF′mx can be estimated from class I rules, the fraction of C′mx that was formed by metabolic conversion of precursor PCB is (1- (BF′mx/AF′mx)), and the fraction of x metabolized to mx and retained in the predator is MF′mx ) (AF′mx - BF′mx)/ M′x. If the AF′mx in the predator is significantly >1.0, metabolic formation in the predator is indicated. In this case, (1-1/AF′mx) is the minimum mx fraction in the predator that was formed by precursor PCB metabolism in the predator, because the metabolite bioaccumulation potential is unlikely to be higher than that of CB153. Another useful parameter that can be calculated is the bioaccumulation/formation efficiency, BFE′predator ) (C′mx) predator/[(C′mx)prey + (MF′mx)predator]. The maximum BFE′ is ca. 1.0, which is when the bioaccumulation efficiency of the metabolite and CB153 are the same and the PCB precursors from the prey are completely biotransformed to methyl sulfone metabolites in the predator. The situation is more complicated for MeSO2-4,4′-DDE since 4,4′-DDE itself is a metabolite of 4,4′-DDT and may also be bioaccumulated or formed at each trophic level.
Results and Discussion The aryl methyl sulfone fraction isolated from polar bear fat (Figure 1A) and ringed seal blubber (Figure 1B) contained structures of persistent MeSO2-PCB congeners consistent with other biota (5-8, 11). That is, the MeSO2-PCBs are (i) 3- and 4-MeSO2-substituted and occur in pairs derived from the same m,p-PCB precursor, (ii) trichloro- to heptachlorosubstituted, and (iii) 2,5-dichloro- or 2,5,6-trichloro-substituted on the MeSO2-containing phenyl ring. In polar bear, all congeners possess at least a 4′-chlorine on the non-MeSO2containing phenyl ring. The minimum requirement of 2,5dichloro-substitution on persistent MeSO2-PCB congeners may be due to the 3,4-epoxide stabilization and thus partial survival from EH-mediated hydrolysis. Either 4′- or 3′,5′chlorine substitution is required in animals with high CYP2Btype activity (birds and terrestrial mammals (19, 22)) to hinder a secondary epoxide formation at the 3′,4′-position of the MeSO2-PCB. Applying these rules to the PCB commercial mixtures that were produced in the highest quantities (Aroclors 1242, 1254, and 1260 and their equivalents manufactured in Europe and Japan (23)), only 12-17 congeners are likely to be biotransformed into MeSO2-PCBs. There are an additional maximum of seven PCBs that could form 3-/ 4-MeSO2-PCBs if PCBs with 2,5-dichloro- or 2,5,6-trichloroand not 4′-chlorine substitution are considered. The 14 3-/ 4-MeSO2-PCB congener pairs identified so far in biota, including those reported here, likely represent the majority of the MeSO2-PCB metabolites formed. MeSO2-PCB or -4,4′-DDE metabolites were below detection (1 ng/g but 10 ng/g, but not exceeding ca. 65 ng/g, and were significantly higher than the 0.02-1.62 ng/g range found in ringed seal blubber (17, 24). The MeSO2-PCB metabolite patterns in seal blubber (Figure 1B) and bear fat (Figure 1A) were similar in composition; however, there were fewer dominant MeSO2PCBs in polar bear. The ∑MeSO2-PCB to ∑PCB ratio in polar bear was only 2-fold higher than ringed seal, despite a 32fold higher level of ∑MeSO2-PCBs (Table 1), and the absence of m,p-PCB precursors. The ∑MeSO2-PCBs in polar bear liver (7) were 4-8-fold higher than in fat. In other biota, MeSO2-PCB and -4,4′-DDE metabolites are known to be preferentially stored in the liver relative to adipose tissue (5, 11, 17). Sources of MeSO2-PCBs and -4,4′-DDE in the Arctic Food Chain. The phase I metabolism of PCBs in mammals is
TABLE 3. Bioaccumulation/Formation Efficiency (BFE′) of 4,4′-DDE and MeSO2-4,4′-DDE in the Polar Bear Food Chain from the Canadian High Arctic BFE′DDEa BFE′mDDEa,d
cod to seal blubberb
seal blubber to bear fatb,c
0.59 ( 0.22 0.004 ( 0.001
0.012 ( 0.007 (0.008) 0.00010 ( 0.00007 (0.005)
a The BFE′ is based on ratios to CB153 (see the Experimental Section). The BFE′ are the arithmetic mean values ( SD. c The number in parentheses is the BFE′ of seal blubber to bear liver. The amount of MeSO2-4,4′-DDE in bear liver was ca. 40% relative to the total amount in bear fat (17). d Only 3-MeSO2-4,4′-DDE was detected in ringed seal and polar bear fat, whereas 2-MeSO2-4,4′-DDE was also detected in polar bear liver (17). b
dependent on the induction and activity of CYP2B-, CYP1A1-, and possibly CYP3A-type isozymes, all of which vary among species (19, 22, 25). The activities of phase II and III enzymatic pathways leading to MeSO2 metabolites are also species variable. In comparison to recalcitrant CB153 (19), the AFs for ∑DDT, 4,4′-DDT and 4,4′-DDE were >1 from cod to seal, indicating high bioaccumulation potential in seal (Table 2). However, significant metabolism in bear was indicated by values
